Distributed Model Predictive Control of Multi-Functional Power Conditioning System for Building Energy Efficiency

This paper presents a model predictive control for a multi-functional power conditioning system. The major useful functions of renewable power conditioning systems in the building power systems are active/reactive power control, harmonic compensation, and unbalance compensation. The multi-functional controller needs appropriate references for multi-functions in order to maximize the utilization of power conditioning system's capacity. The model predictive control is interfaced to inputs of multi-function control by means of individual current references. A minimization cost function is defined by current components with respect to the power losses. This approach will maximize not only PV power harvest but also utilization of the power conditioning system. The expected results are: power quality improvement, energy loss reduction, system reliability improvement, and the enhancement of the life cycle of the building electrical power system. Real time digital simulator based hardware-in-the loop tests and matlab simulation results show that the proposed multi-functional power conditioning system can deliver the active power with minimized reactive power and reduce the harmonics and the unbalance.